JMFit: A SAS Macro for Joint Models of Longitudinal and Survival Data.

Joint models for longitudinal and survival data now have a long history of being used in clinical trials or other studies in which the goal is to assess a treatment effect while accounting for a longitudinal biomarker such as patient-reported outcomes or immune responses. Although software has been developed for fitting the joint model, no software packages are currently available for simultaneously fitting the joint model and assessing the fit of the longitudinal component and the survival component of the model separately as well as the contribution of the longitudinal data to the fit of the survival model. To fulfill this need, we develop a SAS macro, called JMFit. JMFit implements a variety of popular joint models and provides several model assessment measures including the decomposition of AIC and BIC as well as ΔAIC and ΔBIC recently developed in Zhang et al. (2014). Examples with real and simulated data are provided to illustrate the use of JMFit.

Can bundled payment improve quality and efficiency of care for patients with hip fractures?

The current Medicare reimbursement for hip fractures lacks accountability and promotes cost cutting. A bundled payment system-analogous to the Medicare Acute Care Episodes Demonstration for Orthopedic and Cardiovascular Surgery-may help curtail costs, foster communication among health care providers, and improve their accountability for patient outcomes. In hip fracture care, bundled payment may spur development of multidisciplinary best practice guidelines, quality assessment, and reporting, and result in benchmarking and best practices sharing. However, its implementation may face challenges: the need for quality assessment criteria and risk adjustment methods and possible risks of pushing costs outside of Medicare boundaries.

Assessing model fit in joint models of longitudinal and survival data with applications to cancer clinical trials.

Joint models for longitudinal and survival data now have a long history of being used in clinical trials or other studies in which the goal is to assess a treatment effect while accounting for longitudinal assessments such as patient-reported outcomes or tumor response. Compared to using survival data alone, the joint modeling of survival and longitudinal data allows for estimation of direct and indirect treatment effects, thereby resulting in improved efficacy assessment. Although global fit indices such as AIC or BIC can be used to rank joint models, these measures do not provide separate assessments of each component of the joint model. In this paper, we develop a novel decomposition of AIC and BIC (i.e., AIC = AICLong + AICSurv|Long and BIC = BICLong + BICSurv|Long) that allows us to assess the fit of each component of the joint model and in particular to assess the fit of the longitudinal component of the model and the survival component separately. Based on this decomposition, we then propose ΔAICSurv and ΔBICSurv to determine the importance and contribution of the longitudinal data to the model fit of the survival data. Moreover, this decomposition, along with ΔAICSurv and ΔBICSurv, is also quite useful in comparing, for example, trajectory-based joint models and shared parameter joint models and deciding which type of model best fits the survival data. We examine a detailed case study in mesothelioma to apply our proposed methodology along with an extensive set of simulation studies.

Joint modeling of multiple longitudinal patient-reported outcomes and survival.

Researchers often include patient-reported outcomes (PROs) in Phase III clinical trials to demonstrate the value of treatment from the patient's perspective. These data are collected as longitudinal repeated measures and are often censored by occurrence of a clinical event that defines a survival time. Hierarchical Bayesian models having latent individual-level trajectories provide a flexible approach to modeling such multiple outcome types simultaneously. We consider the case of many zeros in the longitudinal data motivating a mixture model, and demonstrate several approaches to modeling multiple longitudinal PROs with survival in a cancer clinical trial. These joint models may enhance Phase III analyses and better inform health care decision makers.

Economic burden of follicular non-Hodgkin's lymphoma.

Follicular non-Hodgkin's lymphoma (FNHL), a slow-growing cancer of the immune system, constitutes about 15-30% of all incident non-Hodgkin's lymphoma in developed countries. Its incidence is rising worldwide. Patients can live many years, but FNHL is considered incurable. We systematically reviewed the English-language MEDLINE-indexed and non-indexed economic literature published in the past 10 years on FNHL, identifying 23 primary economic studies. The economic burden of FNHL is significant, but available data are generally limited to retrospective considerations of hospital-based direct treatment costs, with little information available regarding societal cost of illness. Most direct cost information originates from the US, with one estimate of $US36 000 for the per-patient incremental cost of FNHL care during the first year following diagnosis. The most studied treatment is rituximab, which may offer similar overall costs to fludarabine considering higher resource use with fludarabine complications. Nearly all cost-effectiveness models identified by this review evaluated rituximab for relapsed/refractory FNHL responding to chemotherapy induction. Rituximab is supported as a cost-effective addition to standard chemotherapy by two models in the UK and one in the US, as maintenance therapy instead of stem-cell transplant by one UK model, and as maintenance therapy instead of observation alone by one model each in France, Spain and Canada. The UK National Institute for Health and Clinical Excellence updated guidance on rituximab in February 2008, concluding that it is cost effective when added to induction chemotherapy, and when used as maintenance therapy. No studies of per-patient or national indirect costs of illness were identified, with the only study of indirect costs a Canadian survey documenting lost work productivity. Across all study types identified by our review, the most common focus was on the direct costs of rituximab. As new treatments for FNHL come to market, more real-life cost data are imperative to calculate their relative cost effectiveness.

Bayesian Model Assessment in Joint Modeling of Longitudinal and Survival Data with Applications to Cancer Clinical Trials.

Joint models for longitudinal and survival data are routinely used in clinical trials or other studies to assess a treatment effect while accounting for longitudinal measures such as patient-reported outcomes (PROs). In the Bayesian framework, the deviance information criterion (DIC) and the logarithm of the pseudo marginal likelihood (LPML) are two well-known Bayesian criteria for comparing joint models. However, these criteria do not provide separate assessments of each component of the joint model. In this paper, we develop a novel decomposition of DIC and LPML to assess the fit of the longitudinal and survival components of the joint model, separately. Based on this decomposition, we then propose new Bayesian model assessment criteria, namely, ΔDIC and ΔLPML, to determine the importance and contribution of the longitudinal (survival) data to the model fit of the survival (longitudinal) data. Moreover, we develop an efficient Monte Carlo method for computing the Conditional Predictive Ordinate (CPO) statistics in the joint modeling setting. A simulation study is conducted to examine the empirical performance of the proposed criteria and the proposed methodology is further applied to a case study in mesothelioma.

Cost-effectiveness of first-line induction and maintenance treatment sequences in non-squamous non-small cell lung cancer (NSCLC) in the U.S.

OBJECTIVES: Due to the lack of direct head-to-head trials, there are limited data regarding the comparative effectiveness of induction-maintenance sequences. The objective of this study was to develop a cost-effectiveness model to compare induction-maintenance sequences in the US for the treatment of advanced non-squamous NSCLC. MATERIALS AND METHODS: Decision analytic modelling was used to synthesize the treatment effect and baseline risk estimates for nine induction and maintenance treatment sequences, reflecting treatments used in the US. The model was structured using an area-under-the-curve approach and sensitivity analyses were conducted. Model validation was conducted by an independent third party. RESULTS: All active maintenance therapy-containing regimens, with the exception of gemcitabine+cisplatin (first-line)→erlotinib (maintenance), were more costly than induction-only regimens. Concerning treatments that may be cost effective, the incremental costs per life-year gained were $121,425, $148,994, and $191,270 for gemcitabine+cisplatin→erlotinib versus gemcitabine+cisplatin→best supportive care (BSC), pemetrexed+cisplatin→BSC versus gemcitabine+cisplatin→erlotinib, and for pemetrexed+cisplatin→pemetrexed versus pemetrexed+cisplatin→BSC, respectively. All other regimens were found to be dominated (carboplatin+paclitaxel→BSC; carboplatin+paclitaxel→erlotinib; carboplatin+paclitaxel→pemetrexed; bevacizumab+carboplatin+paclitaxel→bevacizumab) or extendedly dominated (cisplatin+gemcitabine→pemetrexed). Sensitivity analyses demonstrated stability. CONCLUSIONS: Depending on the specific cost-effectiveness threshold used by a decision maker, the most cost-effective treatment sequence may include the referent comparator gemcitabine+cisplatin and the studied regimens of gemcitabine+cisplatin→erlotinib, pemetrexed+cisplatin→BSC, or pemetrexed+cisplatin→pemetrexed.

The economic burden of metastatic breast cancer: a systematic review of literature from developed countries.

OBJECTIVE: Breast cancer, the most common malignant cancer among women in Western countries, has poor prognosis following metastasis. New therapies potentially extend survival, but their value is questioned when benefits are incremental and expensive. The objective of our study was to understand the economic impact of metastatic breast cancer (MBC) and its treatment, and to evaluate the designs of these studies. METHODS: We systematically reviewed the MEDLINE-indexed, English-language literature, identifying 31 articles on the economic evaluation of MBC in 10 developed countries, including studies of per-patient costs, gross national costs, and cost-effectiveness models. We also included health technology assessments (HTAs) from government and regulatory agencies. RESULTS: Total per-patient costs of MBC are only available for Sweden ($17,301-$48,169 annually, depending on patient age (2005 USD)). Most economic analyses of per-patient direct costs originate from the US; across all countries, data indicate that this burden is substantial. Gross national costs of MBC are available only for the UK (cost of incident MBC cases is estimated to be $22 million annually (2002 GBP)). Many cost-effectiveness analyses suggest that a number of new and established treatments are cost-effective compared to standard care in various countries, but many offer small increments in survival. The cost-effectiveness of trastuzumab, capecitabine, and nab-paclitaxel has been evaluated in many recent studies. CONCLUSION: Most economic evaluations of MBC have utilized secondary rather than primary data, and have used scenarios and assumptions which may be inaccurate or outdated. The quality of evidence disseminated to decision-makers could be improved by adherence to best practices in cost-effectiveness analyses.

Validity of 24-h recall ratings of pain severity: biasing effects of "Peak" and "End" pain.

Despite the frequent use of pain recall ratings in clinical research, there remains doubt about the ability of individuals to accurately recall their pain. In particular, previous research indicates the possibility that the most pain experienced during a recall period and the most recent pain experienced (known as peak and end effects, respectively) might bias recall ratings. The current study used data from a published clinical trial to determine the relative validity of a 24-h recall rating of average post-operative pain and the nature and extent of any biasing influence of peak and end effects on nine separate 24-h recall ratings. The results supported a statistically significant but small biasing influence of both peak and end pain. Also, the influence of peak pain was stronger than that of end pain. However, the biasing impact of both peak and end pain together was very small, suggesting that 24-h recall ratings are adequately valid indicants of average pain for patients participating in post-surgery clinical pain trials.

Safety and efficacy of the cyclooxygenase-2 inhibitors parecoxib and valdecoxib after noncardiac surgery.

BACKGROUND: Valdecoxib and its intravenous prodrug parecoxib are reported to increase thromboembolic risk after coronary artery bypass grafting. The authors conducted a randomized trial to examine their safety and analgesic efficacy in patients recovering from major noncardiac surgical procedures. METHODS: The trial was randomized and double-blind, with 10 days of treatment and 30 days of follow-up. Patients (n = 1,062) received either parenteral parecoxib for 3 days and oral valdecoxib for the rest of the treatment period or placebo medications throughout. The frequency of predefined adjudicated postrandomization adverse events, including cardiovascular thromboembolism, renal dysfunction, gastroduodenal ulceration, and wound-healing complications, was assessed in each group. Secondary efficacy endpoints included patients' pain ratings, opioid analgesic consumption (recorded as morphine equivalents), and reports of opioid-related adverse effects. RESULTS: Predefined adjudicated adverse events had similar frequencies among patients who received parecoxib and valdecoxib (2.7%) and placebo patients (3.2%) (P = 0.58), including cardiovascular thromboembolic events (1.0% in each group; P = 1.0). Placebo patients consumed more morphine equivalents (66.2 +/- 92.4 mg) than did patients receiving parecoxib and valdecoxib (43.2 +/- 65.7 mg) (P < 0.001). Placebo patients had higher mean pain ratings on each of study days 2-10 (P < 0.01) and reported more opioid-related symptom distress on days 2-6 (P < 0.01). CONCLUSIONS: Parecoxib and valdecoxib are useful adjuncts to opioids for the treatment of postoperative pain in noncardiac surgical patients. Further study will be required to determine the safety profile of parecoxib and valdecoxib administered to patients with known atherosclerotic disease after noncardiac surgery.